YŞYener Şahin

Yener Şahin

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Yener Şahin has a total of 14 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Validating Large Language Model Judges for Surgical Patient Education: Multi-Expert Study of Reference Standard Reliability and Directional Error (Preprint)

    by Onur Erdoğan, Yener Şahin, Mustafa Sakar, Adnan Dağçınar

    BACKGROUND Large language models (LLMs) are increasingly used to evaluate AI-generated medical content, yet judge validity is commonly inferred from agreement with human ratings without fully characterizing the reliability of the human reference standard or the clinical direction of disagreement. OBJECTIVE This study aimed to evaluate LLM judges for surgical patient education using a validation framework that jointly considers human reference-standard reliability, judge-expert concordance, directional error, and safety-sensitive classification. We additionally examined whether these principles extend to a decision-native probabilistic evaluator. METHODS Five generative LLM judges evaluated 100 craniosynostosis patient-education responses generated by five contestant LLMs (20 questions × 5 generators); three neurosurgeons independently rated all responses on Accuracy, Safety, Empathy, and Completeness. Contestant responses were collected in January 2026, generative-judge evaluations in March 2026, and the Jev-1.13.0 benchmark in October 2026. We examined expert reliability, absolute agreement, directional error, cluster-bootstrap uncertainty, safety classification, self-preference, and a five-judge ensemble. The same benchmark was evaluated separately using Jev-1.13.0, a decision-native probabilistic evaluator that assigns probabilities to predefined ordinal scoring levels and binary clinical decisions, rather than generating single rubric scores and narrative rationales, to examine repeatability, typed safety decisions, calibration, and configuration sensitivity RESULTS Individual expert interrater reliability was limited (ICC[A,1]=0.322; three-expert mean ICC[A,3]=0.587). Generative judge–expert absolute-agreement ICCs ranged from 0.182 to 0.370. Opus-4.6 and GPT-5.4 had negative signed biases (−0.374 and −0.426); Llama-4 Maverick and DeepSeek Reasoner had positive biases (+0.242 and +0.215). A five-judge dimension-median ensemble yielded ICC(A,1)=0.454 but detected 0/5 expert-defined Safety-inadequate responses and automatically accepted 6/10 overall-inadequate responses at the fixed 4.0 cutoff. Jev matched scores yielded ICC(C,1)=0.440, ICC(A,1)=0.141, MAE=0.702 and mean signed error −0.690 against the corrected reference. Its modal Safety decisions detected 4/5 Safety-inadequate responses but falsely flagged 40–42/95 Safety-adequate responses per run. An exploratory minimal-state Jev configuration increased false acceptance from 0/10 to 7/10 in the archived comparison. CONCLUSIONS Agreement alone does not establish the validity of automated medical evaluators. Human reference reliability, signed error, false acceptance, and review burden should be considered alongside aggregate concordance. Ensemble agreement gains need not improve safety detection, and evaluator configuration changes can alter the risk profile. All screening analyses were internal and exploratory, with only five safety-inadequate cases; independent safety-enriched validation is required.

  • Open Access

    Outcomes of Strabismus Surgery in Pediatric Hydrocephalus: Does Shunting Influence Surgical Success?

    by Didem Dizdar Yigit, Sule Nur Kandemir, Volkan Dericioglu, Furkan Cam, Yener Sahin, Eren Cerman, Adnan Dagcinar, Hande Celiker

    Background/Objectives: The primary aim of this study was to evaluate the rate of strabismus and surgical outcomes in pediatric hydrocephalus (HC) patients, and compare the results in patients with and without ventriculoperitoneal (VP) shunt surgery. Methods: This study includes pediatric HC patients who were followed at the pediatric ophthalmology and strabismus unit of a tertiary hospital. Detailed ophthalmological examinations including deviation angle were performed. Preoperative measurements and surgeries were carried out by one strabismus expert. Postoperative deviation of ≤10 prism diopters (PD) was considered successful. Results: In total, 64 subjects were included in this study (29 female, 35 male), and 42 had a VP shunt (shunt+), while 22 were shunt−. The mean age of participants was 8.5 ± 4.6 years, and 50% (n = 32) of them had strabismus (22 esotropia, 10 exotropia). Both groups had similar refractive error and visual acuities (p > 0.05). Amblyopia rates were 21% (9/42) vs. 23% (5/22) in the shunt+ and shunt− groups. In 40% of shunt+ patients (17/42) there was strabismus, compared to 68% of shunt− patients (15/22). Among 22 patients with esotropia, 12 underwent strabismus surgery. Surgical success was achieved in nine patients (75%). In exotropic patients only one had surgery and it was successful. The operation rate in the strabismus group was 41% (13/32). Others mostly did not have surgery due to systemic comorbidities. Patients with strabismus had approximately 3.7 times higher odds of having epilepsy or spina bifida compared to those without strabismus (p = 0.049; OR = 3.69; 95% CI: 1.13–12.11). Conclusions: These findings suggest that children with hydrocephalus may benefit from regular pediatric ophthalmologic follow-up to detect strabismus, amblyopia, and refractive errors at an early stage. Although acceptable alignment outcomes were observed in selected operated patients, the small surgical cohort limits definitive conclusions regarding surgical success or the influence of VP shunting; therefore, treatment decisions should be individualized within a multidisciplinary framework.

  • Modified endoscopic strip craniectomy technique for sagittal craniosynostosis: provides comparable results and avoids bony defects

    by Mustafa Sakar, Serdar Çevik, Semra Işık, Hassan Haidar, Yener Şahin, Bülent Saçak, Yaşar Bayri, Adnan Dağçınar

    Purpose This study describes a modified technique addressing bony defects and incomplete ossification after endoscopic strip craniectomy (ESC) for SC followed by postoperative helmet therapy (PHT). The study aims to delineate quantitative and qualitative outcomes of this modified ESC technique followed by PHT and discern the optimal duration of PHT following ESC. A secondary aim is to address the effects of the technique on bony defects. Methods Patients undergoing ESC followed by PHT between 2017 and 2021 were included. Patient sex, age at surgery, duration of surgery, red blood cell transfusion, length of hospital stay, PHT duration, cephalic index (CI) at multiple time points, and bony defect information were collected. Descriptive and correlative analysis was done. Results Thirty-one patients (25 male, 6 female) were operated in study period. Mean age at surgery was 12.81 weeks, mean duration of surgery was 57.50 min, average transfused RBC volume was 32 cc, mean length of hospital stay was 1.84 days, mean PHT duration was 33.16 weeks, and mean follow-up time was 63.42 weeks. Mean preoperative CI was 70.6, and mean CI at the end of PHT was significantly higher, being 77.1. Maximum improvement in CI (CImax) took place at week 22.97. PHT duration did not have a correlation with CI at last follow up. There were no bony defects. Conclusion Modified ESC technique is effective in successful correction of sagittal craniosynostosis. CImax already takes place, while PHT is continuing, but there is no certain time point for dishelmeting. The technique avoided bony defects and incomplete ossification.